Saturday, June 23, 2007

Anatomy of an alleged academic fraud

I have deleted this post, for now (I may restore it later); however, the images I had posted are still available here.


Rohini Muthuswami said...

There is no way, how much ever good a student is, that gel pictures can turn out to be identical. The biological samples can never ever behave exactly the same on two different occasions, even if you use same gel, same combs, same apparatus.

This affair should have been the eye opener for us in the science field to pay attention to the increasing scientific fraud. Instead we have done what we always do: Sweep it under the carpet and hope that the world will forget all about it.

Unfortunately in science contacts work. For grants and for positions. Therefore, who is going to bell the cat?

Sunil said...

there is just about NO way two different westernblots from two different gels will look identical. The variables are too many. First, the loading of ALL samples have to be exactly the same (and any one working with cell lysates or even with purified proteins knows that it is impossible to get it exactly the same). Second, the exposures of the western blots have to be done in such a way as to get the exact same image. This is impossible. Third, after you scan the image and do change the brightness or sharpness, it makes it even more difficult to exactly match two images.

There is little doubt that the image there was *not kosher* (let us say), and these things usually do come out.

Rahul Siddharthan said...

Scribbler, Sunil: if you are willing to stick your necks out and say this with your real names, it would be very useful. You can email me if you prefer to do that; I will not use your emails without your permission.

(I have already had non-anonymous notes saying that there is no way this can happen by accident, and that is also my opinion: unfortunately, seven of the country's most eminent scientists have signed a report saying otherwise. That is what bothers me, more than the fraud, which -- as Prof Balaram says -- is more commonplace than we care to admit.)

Anonymous said...

Very interesting stuff.I completely agree with Sunil that there is no way in which any two gels can be completely identical,Let me break the different variables for doing western blotting.I will go bottom up,To get an identical bands for western(they use ECL, so will explain this), the films have to be dipped in fixing solution for exactly same time, wahsed exactly same time with water, and developed exactly the same time, even when these processing steps would be identical, the amount of light being emmitted from the gel bands would have to exactly the same number of photons, would have to be exposed exactly for the same duration, upstream, exactly same amount of transfer of the porteins need to be there from gels to membrane paper,and needless to say the proteins on the gels must be exactly the same, this can happen when the exact amount of protein been loaded in the wells,this demands that the protein concentration of the samples out to be accurate enough both times to second place of decimal, and I suppose people working in Cell biology use Bradford assay for protein estimation,and they know how accurate that is!!!!!!Statisticians can help to calculate what is the likelihood of having the exact gels, if we keep a margin of 1-5% error at each step. May God bless 'those 7-members', whom I bet have not run any gel or done western in the last 10 years.
But their reason to cover up is understandable. Most of those seven are SSBs too and so they have a moral obligation to save their kin.This speaks loud and strong as to how they themselves would have got their SSBs too or how they would have been doing their own science.Atleast I personally know a few of those seven, and the way their lab members are pushed to do things.I wish someone would have the courage to look there too.Am sure the answer will be crystal clear then.

Anonymous said...

Why can't some body see what the gist of the paper is and see if it can be verified instead of spending lots of time talking if one blot looks same as another one. If the gist of the paper holds and if it is an important contribution then rest of the matter does not matter much.
Also in these days of world wide web, the importance of publishing in magazines like JBC is not same as before. JBC is not the arbiter of truth. One can publish papers on their own web site and can publish corrections and errata as the need arises. Nothing is set in the stone.People will read your paper if there is a need for it.

Rahul Siddharthan said...

anonymous of July 3: As I said to previous commenters, it would be more useful if you chose to be non-anonymous.

You seem to be saying that data manipulation is OK if your results are correct. Indeed, that was roughly Schön's justification too. I suspect not too many scientists would agree, or at least, not too many would be willing to be quoted as agreeing.

Anonymous said...

from anonymous of July 3 to Rahul:
If a die is cast and some people
refuse to read the dice it is not
the fault of dice.
I would say that Hema and Kundu have
already cast the dice and it is time for their opponents to read the dice.